The performance requirements of semiconductor laser devices have increased over the last few years. As the requirements continue to increase, monolithically integrated devices are increasingly being used. The increased use of monolithically integrated devices require more complex methodologies for examining their performances and for improving the efficiency of the design cycles for such devices.
Conventionally, for a complex device--for example, a monolithic integration of two devices which operate together--difficulty arises in attempting to measure the output performance parameters of the overall device. Uncertainty exists, using conventional methodology, as to which of the two devices are affecting the performance of the overall device. For example, in a monolithically integrated device combining a laser and an expander, if the light outputted from the expander is less than expected, it is difficult to determine if the problem is due to the laser or the expander. For example, in such a device, optical light is expected to be absorbed in the expander. Computer models for predicting the amount of light that should be absorbed are not accurate. Further, for a monolithically integrated device having an expander shaped to allow the beam of light to expand, conventional measuring techniques are incapable of discerning how the beam is transformed as it moves through the device.
It is possible to include a less complex device, such as non-integrated devices including only a laser, to compare to the integrated device performance A deficiency with the present state of the art using a less complex semiconductor laser device as a test device for a more complex integrated device is that unintentional flaws between the device and the test device, such as, for example, bonding damage or process variation across a wafer, are indistinguishable from flaws in the device design. This deficiency is likely to increase with increasing complexity of semiconductor laser devices. Due to the inability to distinguish between flaws in the device design and unintentional process differences between the complex device and the simplified test device, it is difficult to ascribe performance imperfections to the design of the device or in subparts thereof. This lengthens the design cycle time. In addition, it may not be known at the time of device mask design what the optimal test device layout is.
It is therefore necessary to have a design tool and methodology which is capable of accurately measuring the performance parameters of complex devices, thus shortening design cycle time and cutting design costs. Further, it is necessary to have a design tool and methodology which obviates the need for a separate testing device.